Milestone 1: Prototyping

This week, we brainstormed more application areas and decided to change fields entirely, from cooking to shopping. We did some additional observations and brainstormed solutions to problems people encounter while shopping for clothes. We came up with an idea for a kinect-based measurement system, because people need to know their measurements in order to figure out their size, in order to successfully order clothes online or pick the right clothes to try on in the store.

Online Shopping

This storyboard illustrates the main problem we're trying to solve. This person is trying to order a cute dress online, but they have no idea what the size means. We would like it to be a painless process to just stand in front of the Kinect and have your measurements almost magically get taken and spit back out to you. This is far easier and less painful than getting tangled in a measuring tape.

In Person Shopping

This storyboard illustrates the use case for our application in in-person shopping. In the first row, we see the user shopping for jeans, but not sure which size jean he should get. As a result, he gets multiple jeans in each size and brings them all to the dressing room, but the fitting room has a limit on the number of items he can bring in. As a result, he does extra work by carrying around three times as many items as he really needs and is inconvenienced by having to try on his clothes in multiple batches.

In the second row, we see how his experience could be different with our application. Now, when the user is looking to buy jeans, he can quickly find out his measurements using the Kinect application set up in the store. The application will tell him his measurements and suggest the right sized jean to try or buy. As a result, he only needs to carry around one item of clothing.

Concept Art

A visual display will show a few target areas that the user will need to align their bodies against. The user's body will be represented on the screen by point-cloud data, and so the user will be able to align their bodies to the onscreen marks. As the person rotates (full circle? full circle at 90 degree intervals, or just front and side?) their bodies will be scanned and the measurements calculated. The final page will show a list of calculated measurements.

For this initial prototype we went with a metallic, futuristic, body-metric design.

Clothing Measurements

This interface mock-up illustrates one idea we had for how the user could interact with the application to get his or her measurements. The focus of this interface is on the clothing item(s) that the user is interested in buying or making. This way, only the application needs to know which measurements are required for which clothes and the user only has to worry about what item of clothing he or she is interested in.

Once the user selects a clothing item of interest, then the application will proceed to scan the user's body for the relevant measurements.

Continuous Turning

To gather body dimension data, the application needs to know the length and width of the user's body in 3 dimensions. This can be achieved by asking the user to align against pre-determined marks, and then continuously rotate for a giving period of time.

Pros:

  • This approach will capture all the available data
  • It would be awesome to show a rotating point cloud (more fun for user

Cons:

  • Difficult to identify body parts and figure out which points correspond to which places when the user rotates
  • Users may find difficulty in determining the correct rotations speed
  • Processing and storing points throughout the rotation will take a long time

90 Degree Turning

Another way to gather dimension data, is to get the front and side views of the user. There can be 4 snapshots at 0, 90, 180, and 270 degrees. Alternatively only the front (0) and side (90 degrees) views can be used. Each rotation will be prompted by a sound.

Pros:

  • Crisp frames lead to easier identification of body parts
  • Having length and width information for each body part will make interpolating sizes easier
  • User will feel like the process is very scientific and accurate
  • Simple beeps will be clear and non-destracting (people may like prompted 90 degree rotations more than having to figure out a continuous rate of rotation)

Cons:

  • Not using all available data: length & width can be derived from 2D data
  • Results will not be as accurate
  • User will not see as much of the scan in progress (less fun)

Visual/Audio Interface

This storyboard demonstrates that we might not actually need a visual interface. The computer could ask the user to turn, and then either speak or print out the user's dimensions at the end. This may not be completely practical (a person needs to know if they're lined up properly, and visual feedback will probably work best for that, as we notice with bodystorming later), but numerous screens definitely may not be necessary.

Bodystorming

We acted out one way that the product might work. Wendy stood in front of the mirror, centered at the post-it bulls-eye, and turned 90 degrees whenever she heard a "ding" noise. At the end of this process, a screen popped up with her measurements.